Research On Key Technologies Of Received Combining For MIMO Relaying Systems

MIMO relaying system inherits the merits of MIMO and relay technologies,so it can provide extra space freedom to combat fading,increase the transmission reliability and efficiency,and extend the cell coverage,thus having broad prospects in application.In MIMO relaying systems,how to achieve full diversity by making use of the provided space freedom mainly depends on the design of the combined receiving scheme,which requires two key technologies: combining detection and relay selection.The former is utilized to solve the problem of how to recover the data from the signals on different links and antennas effectively and enable the system to attain full diversity,while the latter is adopted to enhance the signal-demodulated efficiency and optimize the utilization of resources without diversity loss when the system contains more than one relays.Therefore,it has academic significance and practical worthiness to research on the combining detection and relay selection for MIMO relaying systems.This dissertation focuses on these key technologies for various kinds of practical MIMO relaying systems.Since spatial multiplexing MIMO relaying systems can provide high transmission rate and diversity gain simultaneously,a maximum likelihood combining(MLC)algorithm is firstly proposed for spatial multiplexing MIMO amplify-and-forward(AF)singlerelay systems in this dissertation.In the proposed MLC algorithm,low-dimension combined signal vector and equivalent channel matrix are derived based on the maximum likelihood(ML)rule,which result in a low computational complexity of the subsequent ML detection.To evaluate the performance of the proposed algorithm,an upper bound on the pairwise error probability(SEP)for the system with MLC algorithm and ML detection is derived.With the help of this bound,we analyze the diversity gain and obtain an upper bound on symbol error probability(SEP).Additionally,the computation complexity for the present and proposed combining detection algorithms are compared.Numerical simulation results show that the proposed MLC algorithm enable the system to achieve the same SEP performance as that of the present algorithm,albeit with a lower complexity.Simulation results also demonstrate the diversity analysis and derived upper bound on the SEP.Thereafter,a weighted combining(WC)algorithm is proposed for spatial multiplexing MIMO decode-and-forward(DF)single-relay systems in the dissertation.The proposed WC algorithm first modifies the near ML detection form by introducing a weighting factor and transforms the detection form into the traditional ML detection form without increasing the dimensions of the combined signal vector and equivalent channel matrix,by estimating the relay detection error vector with its extended complex searching space and the Wirtinger derivative theory.Moreover,the performance loss caused by the estimation of the relay detection vector can be ameliorated by adaptive selecting the value of the weighting factor.Based on the attained approximated upper bound on the PEP,we analyze the diversity gain for the system with the proposed WC algorithm and ML detection and derive an upper bound on the SEP.The computational complexity of different combining detection algorithms for spatial multiplexing MIMO DF single-relay systems are also compared in the dissertation.Numerical simulation results indicate that using the proposed WC algorithm,the system can achieve full diversity and similar SEP performance of using the near ML detector,while has the same complexity order as that of the extended cooperative maximum ratio combining(EC-MRC)algorithm,whose complexity is the lowest.Therefore,it can be inferred that the proposed WC algorithm has a better trade-off between the SEP performance and complexity compared with present algorithms.Furthermore,simulation results validate the theory analysis and derived upper bound on the SEP.Afterwards,this dissertation presents the research on the combining detection and relay selection for the systems extended from the MIMO AF single-relay system.Firstly,two low-complexity combining detection algorithm are proposed for MIMO AF singlerelay aided device-to-device(D2D)systems with the full-rate space-time block code(FSTBC),i.e.,the MLC based and statistic combining part estimation(SCPE)algorithms.The SCPE algorithm can get the statistics from different links’ signal and thereby is friendly to the parallel processing structure.The dissertation also provides the diversity analyses of these two algorithms which are verified by simulations.Numerical simulation results show that the proposed algorithms can achieve the same SEP and only square root complexity as those of the detector based on the ML rule,and compared with the spatial multiplexing MIMO relaying system,MIMO relay aided D2 D system with FSTBC has higher diversity gain.Secondly,a new relay selection algorithm based on the minimum diagonal eigenvalue(MDE)selecting criterion and layer searching is proposed for the spatial multiplexing MIMO AF multiple-relay systems and its computational complexity is theoretically compared with different relay selection algorithms.Simulation results indicate that the SEP performance of the proposed algorithm is much lower than that of the existing relay selection algorithm and is similar to that of the relay selection algorithm using the maximum PEP(MPEP)based selecting criterion and exhaustive searching;as the numbers of relays and antennas increase,the proposed relay selection algorithm consumes the lowest complexity.Finally,the research on the combining detection and relay selection for the systems extended from the MIMO DF single-relay system are present.Firstly,two low-complexity combining detection algorithms,i.e.,the WC based and approximate whitening SCPE(AWSCPE)algorithms,are designed for MIMO DF single-relay aided D2 D systems with FSTBC.Similar to the proposed SCPE algorithm,AWSCPE algorithm is suitable for parallel processing.The difference between these two algorithms is AWSCPE algorithm requires approximate whitening process before the statistic extraction,in order to reduce the influence of relay detection error on the signal demodulation at destination.The diversity analyses for the proposed algorithms are given and validated by numerical results.Simulation results show that with the cost of low complexity,the proposed algorithm can achieve full diversity and approximated SEP performance as that of the near ML rule based detector.Secondly,a multiple error vector estimation(MEVE)based combining detection algorithm is proposed for spatial multiplexing MIMO DF multiple-relay systems.This algorithm is the extension of the WC algorithm in multiple-relay systems,i.e.,when there is only one relay,the MEVE based algorithm is equal to the WC algorithm.The dissertation also presents the MPEP and minimum eigenvalue(MD)based selecting criteria for spatial multiplexing MIMO DF multiple-relay systems and compares the computational complexities of various system schemes using different combining detection algorithms and relay selection algorithms.Thirdly,a universal combining detection algorithm and two relay selecting criteria,i.e.,MPEP and ME based criteria,are proposed for spatial multiplexing MIMO mixed relay systems.Actually,the proposed universal combining detection algorithm is the extension of the MLC algorithm and MEVE based algorithm,i.e.,when the system adopts only one kind relay protocol,the proposed universal combining algorithm becomes the MLC algorithm for AF relaying system or the MEVE based algorithm for DF relaying system.Simulation results demonstrate that adopting the proposed combining detection algorithm and relay selection algorithms enables the spatial multiplexing MIMO DF multiple-relay systems to achieve the near optimal SEP performance and require much less computational complexity than that of the system scheme with the optimal SEP performance.Moreover,this dissertation evaluates the effect of different numbers of relays with various protocols on the SEP performance of the proposed universal combining detection algorithm and relay selection algorithm.